 The reason we started this series is to give a glimpse into the future. Computing is one of the most important discoveries that humanity has ever made and is one of these discoveries that is so transcendental that we've yet to begin to understand where it's going to take us. I often think of computing as not just something like telecommunications or something like writing or something like language. These are all things that have shaped us and shaped humanity in a deep way. I think of computing as the beginning of life itself. I think it's that impactful. Our ability to discover how computation works and embody it in digital systems is like the biological computing that was figured out by proteins and nucleic acids and so on. It's that deep of a discovery and it is going to be shifting and changing our species, the course of life on this planet, the course of potentially the whole galaxy and beyond. So with that kind of cosmic perspective and that level of importance, it's extremely critical in this moment, in this kind of phase transition of this 100 years where we head into things like brain computer interfaces and creating agents with deep intelligence all the way to potentially general intelligence and further on super intelligences. To take a look across the board in all of the areas of computing and try to understand, try to foresee where we're headed. It's critical that we aim these systems and we build these technologies and beings really with a hope towards aiming for the best outcomes for all. That is an extremely vague statement, but unfortunately right now we can't aim for a much more precise structure. As we'll discuss in this series over the week, we'll dive into many different areas of computing and many different lines of thinking across the forefront of many different technologies that when you couple them with computing are going to reshape our society and our civilization and so on. As we think about these possibilities, we need to aim the projects that we're undertaking, trying to mitigate risks as much as possible, especially existential risks, for example a rogue AGI that is unaligned with the species and unaligned with all life, and try to aim for extremely beneficial outcomes for everybody, for all of humanity, for all living systems, and for all beings. We'll have to contend in this century with extremely deep hard questions like what does it mean to boot up a new life form? What does it mean to boot up a new intelligence at similar levels as us? What does it mean to interconnect humans with wires to the internet? What does it mean to interconnect humans to each other so they can think together? That's in our lifetimes going to be the range of challenges that we're going to have to figure out. So what this series is about is to help us think about the short term in context of that long term. How are the things developing right now in the last few years, in the next five, in the next 10, in the next 15, and 20 years, and what are the kinds of projects and things that we should be focusing on so that we can aim for the best outcomes possible? So we're going to be talking about important short to midterm things that we can all work on and do with the context of the long term beneficial outcomes. With that I'm really excited to welcome Allison to be our first speaker. Allison is somebody that I've grown to admire very deeply. Allison is an extremely thoughtful and engaging thinker really, a thoughtful thinker. She's an extremely thoughtful, engaging person who has been working extremely hard to help many communities unlock their potential in a super deep way. She leads now the Foresight Institute, which is an institution that has been helping us understand the future of technologies like nanotechnology, biotech, AI systems, and more. And Allison has really transformed the entire organization in the community and has built something truly amazing and special. Not only that, but she's pioneered a whole set of seminars and discussions that you can join on YouTube that are all out there for you with some of the most important technology developers of our time. And these are extremely accessible conversations that take the extremely deep complex subjects into something that you can understand the implications of. And Allison is extremely, extremely good at helping tease out the important things that we should take away and what things we should work on. So she's been a phenomenal connector for these broader communities. And I think she's one of the people on the planet best equipped to help us think through these future questions. And I think if in 100 years humanity is still around, it'll be in significant part thanks to Allison's great work over the last few years and the next. So with that, we'll hear a format of the event. So after this short introduction, Allison will give a presentation and then after that, we'll be in a fireside chat. So I'm here in Lisbon. Allison will join us remotely. And then after that, so we'll have a fireside chat. And after that, we'll be able to hear questions from people both in person and remote. So here in person, we have a mic up here. So please think of questions as the presentation goes on and as the fireside goes on. At the end, we will call for questions. So please line up there. And if you're kind of in the middle sections, there's somebody running around with a mic. But please, if you're sitting on the edge, please stand up and go to the mic. And remotely, if you want to ask a question, there's going to be a, you can either ask a question directly on the live stream tweet. So you can kind of subtweet there or use a hashtag. I don't know what the hashtag of the event is, but I'll ask and figure out before we open up for Q&A. It's PL Breakthroughs. Great. So PL Breakthroughs is the hashtag. You can ask a question with that. We'll be monitoring Twitter and here at. And after hearing this introduction, if you know somebody that would love to be part of this conversation, especially live, to be able to ask questions, send them a link to the live stream so they can catch it. And of course, separately, this whole thing is going to be recorded and stored for posterity. So we will be able to ask it there too. Thank you. All right. Hi. Can you guys hear me? I'm assuming it's a yes. If not, please, someone do something about it. Yeah. Well, thank you. I really have to kind of start by thanking all of you guys for coming. I can't see you, but I certainly appreciate that many of you have made it out. What I'll try to do because, you know, to keep it also interesting and enticing for you, I'll keep my talk as brief as I can. So it'll be very fast with the interest of giving you kind of a menu of options to choose from later to dive into, into a longer Q&A. So if something piques your interest, I really welcome you guys to just point it out afterwards. So my name is Alison. I'm from Forsyte Institute. And I run that organization now, have been with it for 10 years, but the organization has existed since 1986. So that's a really, really, really long time ago in terms of the technology and technological progress that you see. And what really inspired me to come to the org is that since the early days, really before many of these technologies were actually possible, people have been dreaming really big and ambitiously about the future and have been trying to really strive to create technological change across different domains. And so I'm, like, really thrilled that, you know, I can stand on the shoulders of many folks that have worked in such a long time to build out a broader community in that space. And so what we do is we try to support early stage science and tech development that is a bit too ambitious, too interdisciplinary, perhaps, you know, too early stage for legacy funders to support. And we have do so through a variety of different means. I can go back to them later, but it's prizes, fellowships, workshops, seminars, and so forth. And so we really try to kindle an ecosystem where there isn't one yet. And so I wouldn't say that we're like the best at any individual, at any individual technological area. But I think what we can do is like kind of like give people a lay of the land of what may be out there in other technological areas that they may not know about yet that will be very important for their fields a little bit later on. And so what I want to do today, and I'm hoping that screen share and everything works. If not, I will just rely on you guys to tell me. But what I will do today is give you a quick run through really of the main areas that we focus on, which are these biotech and health extension, neuro tech, molecular machines, intelligent cooperation. And I'll talk a little bit about what they are. And I want to show you a bit how computing speeds are progress in any one of these areas, because I think as you know, all of you know, it's certainly one of these technologies that isn't just, you know, one inch technology in one area, but that really means a lot for the development of other technologies across the board. So that's what I'll do. And in the interest of time, I will just gloss over a few of top kind of like seminar highlights. And if you want me to double click on them, we can certainly do so. Okay, so let's start with biotech and health extension, like as a broad goal, really this group, but this technical group and seminar group and workshop series that we have as trying to do is really increase the quantity and quality of life for individuals currently alive and alive in the future. And so there's a few real highlights, I think that that we've had this year. And so this is Morgan Levine with Jamie Justice. And what they're doing is that they're developing, they're using computational biology for really developing new biomarker standardizations to really measure human health across the board. And so really the idea here is that in longevity, you cannot tackle what you cannot measure, but there's so many different aging clocks. And so it would be great if at first and foremost, of all, we had somewhat of a standardization of like, what is it actually that we're trying to improve here. And so I think really it starts there and they have done fantastic work. Morgan was back then still at Yale, now she's at Alton's Labs. And the work is really like really, really mind blowing. And I encourage you to check that out, but that certainly wouldn't have been possible without many of the computational biology tools that we have today. Second one here, and perhaps a little bit further out already, is George Church's work, which I'm sure that many of you guys know. And he basically takes the parallels between how the body encodes information with nucleotides and how computers encode information with binary zeros and ones, and suggests that maybe that parallel means that we can eventually build synthetic biological systems in the same way and as creatively as we currently build digital systems. And so he starts really with resistance to all viruses, goes to enhance organs, multiplex editing, and age reversal, which is something that this group cares about a lot. But you know, he's certainly done a lot of work. I don't know any other university professor that has launched that many companies and advice so many. But I think many of this would really not be possible without much of the computing progress that we currently have. This one is a little bit further out still, Michael Levin, which I think that some of you at least know, not everyone does know him. So here's this new concept of really bioelectricity and explores how we can create xenobots. And so those are really like newly generated small self healing biological machines that I created from cells that can move around, they can push a payload, they can retain memory, which is a newer version that they created. And crucially, they can self heal. So here you really have this concept of regeneration. As you can see here, he is creating really through the stimulation little froglet legs again on this on this on this emerging front. And ideally, and I think like kind of crazily also they can these xenobots like these kind of like newly created biological robots can also exhibit collective behavior in the presence of other bots. So they can really have this kind of swarm like intelligence already. And so he's kind of like really kind of just combining a variety of different fields into something that, you know, is really still pretty unheard of to as far as I was concerned before I watched I watched this seminar. Next one up, if we take the next track that we that we focus on and foster, which is newer tag here, I will keep this very brief because we will hear from Adam Arbelstone and from Nick Bostrom, who will be discussing newer tag a little bit later in the series, and I certainly cannot help to do as much of a good job as they can. But just to give you like a few early tools on the goal towards brain computer interfaces, perhaps later, whole brain and relations is that at Boyden whom I'm sure is also named that some of you know, but he's at MIT. And he basically developed really like a few new novel technologies to help him really do his work. So he's trying to create maps and real models of the brain circuitry. And so he's using optogenetics and that means that you can already control neuronal activity with light, which is something, you know, that is fairly new. And another bit that many of you may be heard of in his TED talk is that he developed this mechanism of expansion microscopy. So you can basically expand nanoscale specimens by literally using a gel that swells in water. So you can have these specimens and become much larger over time and actually like look at the properties that you have going on here. And so here he really shows just like how small the neuronal activity is that we need to kind of like be able to figure out to really increase our performance on the brain. Next one up, this one is one that I think is interesting. It's pretty new term. It's already exists. In fact, it's a company from David Eagleman at Stanford. And so this is a person who is deaf. And then he is wearing a vest that David and his team developed that also exists as a wristband now. And with this vest and wristband, he someone that's a square student can speak words at him and he then hears them through the vibrations in the vest and can write them down on the board. And so it's basically this new way of translating basically outside information into something that into vibrations that the brain can then make sense of, certainly not this, you know, would really have gotten have been possible without much of the computing progress that we see already. If we take the next goal of molecular machines here, we're getting really into the terrain that force that was founded on. So we were founded by Eric Drexler and Christine Peterson on the book Engine of Creation. And so that envisions this long term future of what would it mean if we can actually create things from the bottom up with atomic precision. So really by moving potentially eventually individual atoms around. So that's a very long term goal. But it would mean that you can create almost anything incredibly precisely cheap without without waste. And you can create entirely new physical objects. And so we're not there yet by far, but we have made an insane amount of progress over the past few years. And so I just want to, you know, briefly give you a quick walkthrough of what's already impossible here. And so this will be brief, but short and sweet, because I think most of the progress we did that I was blown away by this year came from that group. And so here we have that's coming out of Baker's lab at University of Washington. Here's Alexis Kobe. And he basically asked, what can we hallucinate de novo proteins already? So this is, you know, the the the dream tools that many of us have played around with for machine learning. And what you can potentially do is using similar tools to really help with protein folding on the path to creating atomic precise machines using proteins as individual components. They've published a few incredible papers on this work. And yeah, it's really pure magic what's coming out of the lab recently. And that is really using kind of state of the art Rosetta tools, and others that come from the machine learning community now. This one is a little bit more out there because it involves a physical component too. So that's Lee Cronin at the University of Glasgow. And he is basically building a computer. And so it's basically software that can translate the words that a chemist speaks into recipes for molecules that a robot can computer can then understand and produce. So this is this non discovery robot. It's a physical device. And it's really linked to kind of like end to end to this tool by which, you know, chemists can like try to communicate with it already a little bit. Like you speak to, you know, like to to an assistant, basically. And so it's really cool work. It's like still advancing, but maybe one level up. And this is Adam Marlestone, who's I think speaking on Friday, if I'm not correct on Thursday to you guys, so getting excited. He definitely has his hands in many different individual fields. But here he's kind of taking up this more longer term goal of a molecular 3d printer. And so he's asking, what could it look like if we did 3d printing with Lego like macro molecular building blocks and use self assembly and motors that are based on DNA or protein origami. And so DNA origami is this new field that was buried by will she and a few others and at Harvard and other organizations. And he's basically trying to see what would it look like if we take many of the individual components and pieces and put them together that people are already working on. And so this is his concept of molecular additive manufacturing. And so he's really proposing to create a molecular 3d printer that can print individual bits with a greater and greater precision. So here we're getting much closer to the early goals of molecular and nanotechnology. And it's it's really great work. Next one up, we have something that is has really blown me away. So this is a presentation given at a workshop that we had this year. And it's really quite something. So basically it's here they're proposing to simulate small proteins by using a building, a physical building as a computer. And so you basically see people here using tables as computing surfaces, physical objects, such as notepads and so forth, that are captured by overhead cameras as the interface and literally creating these molecules, moving them about with laser pointers and kind of playing around with each other's molecules, but physically. And so this is this kind of like idea that if we do the physicality, you have this kind of like shared mental mental map of what other people are working on. Brad Victor at Dynamic Land, maybe a name for some of you doing is doing absolutely mind boggling work and has here partnered up with Sean Douglas. And you can look at this basically, you can go to the lab and try it out. It's pretty, it's pretty fantastical. And then, you know, here we come back to the more long-term goals of molecular nanotechnology. There was also at one of our workshops this year. And that's Eric Drexler who founded Foresight. He's now at FHI. And so he's basically thinking like, can we use massive multiplayer online gaming for complex molecular systems design? So could we create like using a game engine like Godot, basically a system that is interesting enough for a bunch of people to pile in on an online game using real data and helping people to create these more complex machines in a way that is fun. And so he's trying to outsource much of the work that people are currently doing in the lab. And it's a project that is now under development. And so I really encourage you guys to stay up to speed on what they're working on. Okay, cool. So next one up, we have Human to Human Corporation. And so this is something where many of you will feel at home. I will show a few familiar faces here. The goal here really of this group is to increase, first of all, human to human corporations. So what could it mean to really unlock our potential? This one is the name that I'm hoping most of you will recognize by now. And so Juan Mune really joined us to discuss this notion of Pareto-Topia mechanism design. So what would it look like if we could increase the human ability to engage in Pareto preferred interactions over time? And the way we can do that is using software already to reshape really much of the mechanisms that human civilizations run on. And some of them include law like contracts, rules, rights and courts. And so you're all deeply, internally familiar with blockchain smart contracts. So that's that bit. And then economics, such as money, property, stocks, derivatives. And I think, you know, this, I don't have to spend much time on you, you can block that, that one good, I think. But I think the kind of amount of potential, as Juan shows here, right, that can be unlocked really through like just coming up with better mechanisms for human cooperation is absolutely, is kind of, is kind of stupefying. Next one in a similar vein here is where Baladu joined us to discuss really the concept of the network state. And so the idea here really is that, you know, we already know how to start our own company, we know how to start our own community, we know how to start our own currency, but now soon enough, we'll also be able to start really our own soventies. And he, you know, came out with this concept a little while ago and shaped much of the discourse really up to now to the extent that people are already moving into, or at least like using many of their ideas to now prototype what previously we're called C-STEDs, or we're called charter cities. Now you can build them kind of like on the network and move them to potentially physical locations. That is also possible because of computing, to some extent. Next up, we have much of the field that, you know, you will probably hear about a lot about this week from wonderful folks that are already present. And that is really this concept of creating DAOs to send less autonomous organizations. And such as, you know, here, Vida-DAO, they joined us, I think it was two years ago back then to really like, you know, prototype this idea. And I think now, as you can see from the exploding d-size space, it's really kind of, it's really making headways. And even that, you know, wouldn't have been possible without much of the compute that we have today. Next one up, we have this notion of more collective intelligence. So many of you guys, I think, you know, already are pretty good at making good guesses about what will eventually become important in the future. But Anthony Aguirre here from Metaculous takes it further and asks, like, what do brains, markets, and science really have in common? It's this kind of like, really idea of like collective sense making in a pretty rigorous way, and uses a few tools that are being built. For example, I already encourage you to check out Metaculous. I certainly prefer it over my daily news, daily news digest, and it's a forecasting platform that has built up in a really great community for predicting and for forecasting the long-term future. And it's already possible, and you can really use a few statistic analysis tools over those predictions to get much, much better results over time. But, you know, I think the better we get at tracking wisdom, but also accountability of individuals, I think the more we can unleash on a general scale. So this is a little bit about what we can do for human to human cooperation. Now, if we look at human AI cooperation, which is something that I think we will figure, we will have to figure out if we want to have any future at all, it's something that Nick Bustin will talk about a lot tomorrow. So I keep it super brief and just name three bits. One is I think one thing that we really need to figure out if we want to have any positive future at all is computer security. And so I think computing really can't go well on the very long term without having systems that are probably secure and that can withstand good red teams. And so here we have Donut Heiser from SEL4, which is a probably secure microkernel. And I think, crucially, it's the only one where a DARPA red team couldn't break in. They didn't just not make and they didn't just they weren't just not able to break in, but they also weren't able to make any progress in breaking in after the entirety of the time that they were allowed to have to use to do the red team. So it's a really, really secure system from the University of Sydney. And it is open source design. So it's really cool. And, yeah, Donut and the team is absolutely amazing. So I think that is kind of like a number one layer we need to have. And from there we can, I think, think a little bit more broader. So this one is George's Casey's from his leading research, I think still at the open mind. And he's basically discussing how cryptography distributed computing and differential privacy can unlock access to orders of magnitude more data. So for example, in particular fields as medicine, it's really important that like much of the sensitive data really doesn't get leaked. And through that also much of the bits that we would like to do already with or we could do really to improve people's lives with the sensitive information is currently still blocked to us. And so perhaps there are mathematically secure ways in which we can unlock some of these like data silos to create kind of like, yeah, to really compute on that on metadata to, to really improve the lives of many more using using data that is currently still locked up. We just had a longer workshop on security cryptography and AI, and that really explored kind of this path of just what is the intersection between cryptography and machine learning? And what can we learn from that? Are there any undervalued areas a bit more? So I'm happy to dive into that a little bit later. If you guys want to in the Q&A. But this one is the final one in this, in this kind of like, yeah, in this kind of like speed run. And that is the universal constructor. And so this is Chiara Maletto who works with David Deutsch at the University of Oxford. And she is basically saying we should go beyond the universal computer. She's basically saying the universal computer wasn't quite broad enough. What we actually want is a programmable machine that can perform any task that is physically allowed in the physical world, right? Which computers cannot do right now, including creating a copy of itself. So like a physical copy of itself. And so this is, you know, obviously quite a bold long-term move, but she's basically saying that we should think a little bit broader and more ambitiously even about like what computing can realize in the physical domain. This is still, you know, a pretty abstract universal constructor. But I think as I've shown you with some of the work in the molecular machines group, we're getting closer to these kind of like 3D printing and physical worlds realities through computing really. And so yeah, the limitations are really like almost not given here. There's a pretty endless horizon if we can do that. So in a nutshell, I've said very, very briefly a few areas of how computing speeds of progress in different technological areas. One thing I do want to just briefly mention, which I think is really, really important. And I hope we get to it in the Q&A, is the fact that computing also speeds up risks. So all of these areas come with a variety of different risks. And so I think there's two types of risks really that we need to learn how to deal with. And I want to bring them out here because I want to appeal to you guys to help us solve them. One is the risks that is pretty obvious, like, and that's the risk of small kills all. It's a term, I think, coined by Bostrom. And so he's basically showing really, and as you can see even from this, like on the second thought, technological proliferation, especially through computing technologies, can allow a small number of actors. So like even just civilians to cause a larger and larger destruction. So a few things that come to mind here are bio weapons, right, that you can really just design in your basement eventually. It's killer bots, like physical robotics that you can print potentially. And it's cyber attacks that, you know, are already causing massive financial loss, but could also really lead to the destruction of much of the civilization infrastructure that we rely on. So that's really bad. And then I think in the face of that, usually the solutions we come up with also create another risk on the other hand. And so that is this part of single point of failure. And so I think that the general solutions that are discussed to these problems also involve centralized actors often that have really large surveillance capabilities. So like, what would it mean if we have really one actor that could surveil kind of like the development of these technologies and can really enforce counteractions when we don't like what's going on? But I think these types of solutions, you know, that we currently just come to mind or like bubble up immediately as a default solution come with their own risks. And so on the one hand, you have power abuse. I mean, we all know the notion of power corrupts and absolute power corrupts. Absolutely. So I think that really creates a single point of failure. They're vulnerable to external attack and coercion. And also, you know, usually if you have a centralized system, it is just not that easy to adapt. But it may lead to some stagnation on the long run where we're just not able to keep up with the risks that we're creating. And so that's why I'm appealing to you guys. You are a member of the decentralized crypto community. And so maybe you can build a third path of decentralized systems that can account for both of these things. On the one hand, avoid risks from technological collaboration without creating these single points of failures on the other end. We propose a few of these solutions in a new book that's out. It's called Gaming in the Future and it's kind of exploring the long term potential of crypto technologies. It's now up for pre-order on our website. So I encourage you to check it out if you'd like. And we make a few very concrete suggestions here. I think that this is really something where you guys in this particular community that I think I'm speaking to now can really have a really large impact. I think thinking a bit more longer term about how can crypto technologies give us secure, free and high-tech futures but that are nevertheless flourishing. And if you want to get involved, finally, there's like I think three ways in doing so. On the one hand, yes, you can absolutely join foresight. There's many different ways. We have a variety of different technology groups. That's the seminars I just showed you. All of these come with their own workshops and reports. So I've just often shown you like a few of these workshop videos and so forth. You can check them out if you want. You can apply to join them. We have prizes and fellowships in these areas. We've given out the five-month price since 1993. We just launched together with Mieterdal and a few other wonderful partners. The Long Long Jeopardy Prize. We have fellowships in all of our different technical tracks and we have coming up Vision Weekend in Europe and the US, which is our annual member gathering where we kind of bring together top folks across different tracks in Europe and the US. So I do invite you to just poke around the website. There's a ton of things you can check out and you can get involved in these seminar groups that have shown you bits and pieces from all virtual groups. So you can not only watch the videos, but you can actually apply to join them if you are a technical person in that field. If you want to get an overview of a field and actually dive in and see where you can make progress on, we are currently building tech trees and so I invite you to just check out one of these trees. We are listing basically the main areas that people are working on in these fields, including individual labs and so forth, that you can join and including trying to make a few connections between different areas and kind of like showing people that are new to the field where they can have the biggest impact on the part of the long-term progress in these fields. So if you're interested in like really getting an overview of the field, I encourage you to check out those. And finally, if you want to work on the risks, I encourage you to work on the crypto tech that I just talked about in the Game in the Future book, but also really there's this other website that we've created called existentialhope.com and here we basically list positive and negative futures across all of these domains with the idea that we inspire people for the positive worlds and we basically also immediately list a few organizations really that you can work on in individual areas that are across the board that are beyond foresight much more and so this is this more like positive futures hub and so if you are larger on a larger scale interested in getting involved, I encourage you to check out that website too. I can speak much more in depth about any of those things, but if you have questions just tweet at me, my DMs are open and yeah, I'm super excited to be here. Thank you so, so much, Juan, for organizing this wonderful series with such an incredible amount of bright minds including all of you guys and I'm really excited to dig into a few more of these areas in the Q&A. So thanks a lot for having me. It was really wonderful to put this together and very excited to be here and I'm hoping that yeah, together we can really make some progress. It's already a lot has happened and yet we still have some way to go. Thank you Allison and feel free to if you want to especially in the middle as we ask about some of these, if you feel like sharing your screen again, maybe feel free to do that. But some here, let me know if you can see me. Good? Yeah, great. And so we'll have a little bit of lag and so I'll try speaking until you like stop and then I'll give you a little bit of time because I don't want to kind of interrupt you. Thank you so much. That was an amazing survey and overview and one of the things that you do extremely well is help everyone connect with these extremely deep fields. So thank you, thank you for that. I wanted to kind of start talking about you a little bit and foresight and then dig into these specific, these four topics and four areas, walking through how you think they might develop over the next, you know, five, 10, 15, 20 years and maybe we can stop and think about some of the kind of like really exciting, optimistic things that might kind of pick up speed like which of these things is like you know potentially going to gather a lot of steam quickly or which are like you think are like significant risks that you don't think are being addressed properly that maybe some of us can go and work on. And then after that we'll kind of like think about the future together and we'll think about kind of like where we're headed. So let's start with you. One of the things that I like hearing from super brilliant people like you is how did you come to be you? Like what are the kinds of works or experiences or like things you learned or thought about that helped shape you into who you are today? Yeah well first I think I want to say that you know much of the fields that I've just glossed over I think is I have an ability to just have enough of a grog to ask the kind of super questions that then make people explain things in a way that and that is actually useful for an outsider that may want to develop skills in that. So you know I think one bit that I certainly enjoy is kind of like weaving a little bit more of the connections across these fields and figuring out you know if you talk to people in these fields they are so busy like especially in academia like you know they have so much on their plate there's just not time to look beyond the immediate horizon of your in not even your discipline your sub-discipline right your department and so what we're trying to do is a little bit like kind of weave the web and try to show people what's like just beyond the event horizon and I think that is just really like so pretty undervalued like people just don't have the time to do that and I certainly enjoy doing that a lot so my background is in philosophy and so I think you know like having a little bit more of a border angle has been always like pretty exciting to me I do want to say like on a personal note I guess I just never really like the idea of death very much and so you know like I think growing up it was kind of like untenable to me to you know have to wither away eventually and see everyone I love around me die and wither and slowly decay and then I think as you kind of like go out of puberty you realize that there really are no adults in the larger room of civilization to the extent that it is not even true that civilization will survive that is just not a thought you have as a child like as a kid you just think that history has existed since forever in a day and all you have to do that all you have to do is kind of like stick around with it to continue to live and then over time you really realize just that the technologies that we're creating and like the general geopolitical conflicts that we're having don't even allow for us to really be secure that civilization will survive and so I think these like tooth like kind of existential worries about my own mortality and about civilizations mortality but then also paired with this like really insane potential brought me to Fawcite in the sense that I think yeah they're really one of the more optimistic organizations I knew that were deeply tech native without being pollyannish about many of the risks that were developed and one thing that you know I think people do really well here is that they have this more approach of like differential tech development that is not developed as I think it's a term developed by open fill but that you know is really trying to look at the risks that technologies are creating as they as tech is progressing but built tech that can address these risks and like kind of like build the safer and more secure tech first and so having this really I think yeah kind of like smart strategic approach to tech development that is that is pretty attractive I think to me so that's a brief answer yeah I also vividly remember when I was a kid I learned that like people died and it was like what are you kidding me so yeah we all have to like learn that and I think in college I was kind of finally becoming hopeful that we could actually be one of the last generations of ideally one of the first generations to survive but in the worst case one of the last generations to to not make it so one kind of following question about about you especially for people who are developing their their their work and technology and so on what is some advice that you have for you know people like you mentioned that are kind of deep into their field and are having trouble not just keeping up with what's changing in their field but are but need to kind of look up and reorient to see what other really amazing things are like naturally of course like be able to explore some of the really amazing synthesizing information that you and and others are putting putting together but what are some of the things like individuals should kind of do to make sure that like their deep work is on the most important problem like how would you point them towards like areas of work yeah well it's a difficult question because yeah you don't know what you don't know really um but um one thing at least that we've often done for building these for building other tech trees we've had a really bottom up approach so we ask people always on the one hand if you look at your individual fields what is an enabling technology so for example like what would it look like um you know to like what really led into your field like you know what was super necessary sorry for the thing that you currently work on to exist and then on the other hand like what is like what a potential downstream applications and so once you have that you already know kind of like some of the neighboring communities at least um that you know are of interest to you if you want to push progress forward on the one hand you can see and go to the enabling tech developers and see do they have better tools than I know about like you know there's probably a lot of progress has happened since last time I checked and on the other hand you can build and go to the ones of the immediate downstream applications that you're already aware of and just tell them like look hey I have this thing I'm working on um you know is there already something like an application down the road one thing that has been incredibly hope inducing for me is we have a fellowship and we had our review process the past few weeks and um we've had over 200 applications and many of them are actually in their teens and I didn't know that they're in their teens because we have like a blinded review right and then you find out eventually and it's absolutely mind boggling what young people are already working on um and then it's also really mind boggling the second thing is that many people in academia kind of already have a company but they may not know about it so for you guys in academia I would say talk to people that are working in the application area that you think you're working towards likely is that many of the companies that have already out there have been started on less tech than you already have in your lab you just need to find the right people um that can really help you speed things up and move things out there into the world I'm happy to if you think that you have a hunch of something that you're working on like if you're ready to reach out um but but um but yeah I think those are like definitely things that that are maybe a little bit more immediately practically irrelevant for someone yeah it's a good thought access yeah one of the things that I see um preventing and stopping a lot of people from interacting and connecting across those uh those communities and reaching out to people in different um labs or different companies and so on is this uh short-term um perspective around getting scooped like so or running into an IP problem so a lot of researchers don't collaborate because they feel like if they collaborate then somebody else might get the paper and if they get the paper they get the reputation and they get their reputation they get the money they get the money they do the more work and so it is just plunged all of our brightest minds into this like highly conservative environment that doesn't share and that happens also with IP uh when people develop some technology and are starting to build something um they get into significant problems potentially trying to um collaborate with others because there's all these kind of like IP roadblocks in those cases people might know about the approach but they don't they can't work together um how do you think about this and and do you think like there's um you know hey how do you feel about like these incentive structures and systems and um do you have like thoughts on like how we should reshape them oh yeah i guess that you know like um the whole kind of recent um wonderful movements of like dci i think is really trying to do much of that um i think it's it's released the early stage in those bits um i think i mean look there's a few i think exciting efforts i just want to draw attention to they're not all dci but one of them is papa so basically ben reinhardt um is creating this kind of like private alpha uh organization here in the u.s. adam marble stone will join you guys in a few days um he's creating uh fRO so those are focused research organizations that are basically combining this kind of startup thinking um with more long-term technological goals so for example what i showed you and molecular machines like a totally precise manufacturing we won't get there in a lab like that is not a thing a lab can do that requires a very very directed very interdisciplinary approach of a bunch of people a ton of resources ton of money a long timeline for people to work on and so he's trying to create these kind of um focused research organizations that are run like companies um but nevertheless have long-term kind of scientific goals that can push us further um recently i think i've met people from aria and jedi which are european versions of the us opera and they're new who knew that um we didn't really have them in europe before that but anyway now in the uk and europe we at least have an opera so that's cool um and so they are doing really really really great work i encourage you to check them out jelly and aria respectively and then there's a bunch of like individual i think you know like honestly frankly um kind of yeah i hide it with individuals that have like high hopes and dreams for the future that fund entire ecosystems right now and i think it's the more inspiring thing that i've seen in a pretty long time this is has been really unheard of like even in the last 10 years it has existed to some extent but now people are just really taking the future into their own hands but i'm putting the money with them out of this i found fantastic organizations and are just really i think developing lots of progress so that's i think happening more in the normal i guess you know more traditional sectors and then the whole d-size space obviously and i'm sure that you will discuss this more like uh in individual sessions too i'm sure that you know like you certainly have a really good eye on the ball here as well but like this notion that we can finally use these tools in uh in cryptography uh in in the underlying parts of of much of the crypto sector and to kind of like reshape the way that scientific funding and incentivization can happen is i think really interesting molecule obviously has the notion of an ip nft uh and what you can unlock uh through uh through actually being able to fractionalize uh ip ownership upfront um and then there's you know a host of other i think really interesting uh really interesting innovations in that space one thing for example you know that could be a long time application of the tech tree that i showed earlier like is imagine you had this kind of technology tree of an entire area right in trend economy talked about it um in a really great post i imagine you had a tech tree that was just mapping out how do we get from here to bring computer interfaces which is something that you know we currently hoping to do um and then what would it look like if you can actually uh put individual bounties on solving specific notes and over the long time those are not just only bounties but they can actually be like syndicate doubts so for example you can lock up the capital and say hey i want to solve this particular note in bio preservation for cryonics on the longevity tree um i'm going to put my money here i will have uh that money go out um to the best proposal that goes uh that gets proposed for solving this note it can be a multi-sig whatever you want but then ideally that money kind of transforms into initial seed funding for an entire note and so you basically have this kind of like a tech tree serving as an api or the call for project proposals where people can just propose more longer term bets and it's a mix between a philanthropic funding mechanism or an investment mechanism depending on like really how you play it if it's a bounty or if it's a syndicate doubt and so you have all of these kind of like new mechanism design opportunities now that i think are super super inspiring yeah all the tech trees uh model is so so exciting um and by the way if anybody makes games out there and makes tech trees uh come talk to us because uh if we can make these tech trees extremely understandable and and amazing from a ux perspective uh it'll help civilization on a massive scale and so like really like help us make it fun make it make it interesting um uh it is a super valuable valuable resource um how do you think about ip and ip structure like in my perspective ip is mostly um you know in in software at least it's just primarily seen as a way to delay and slow down and block other groups uh in a pretty negative way i think most of the software industry tends to look at uh at patents very very negatively um in biotech ip is really the only way in which small groups can um you know how labs can turn their things into um a small uh you know kind of like drug discovery company that can then later get acquired by like a large pharma and then um scaled now we're starting to see some small um small of those biotechs that now want to become platform technologies that are where maybe the asset becomes a small company that gets sold but like they themselves remain to be able to kind of become like a factory of more assets uh when they have some kind of like platform technology that they can like you know figure out more more of these um it seems here like it's one of these areas that is ripe for changing the incentive structure with like a small adjustment to the marketplace uh that can radically change the the rate of innovation um i like to think about this kind of like airbnb so before airbnb uh all the rooms were there the people were there the people were traveling it's just they couldn't interact and they couldn't transact because the marketplace was like totally broken like you had to ask for people um ask for friends and so on and and try to like find a room to crash in and so on um but all of the resources were there they just needed to be wired up into a better marketplace and so whatever airbnb did is like it changed the structure of the transaction it cleaned up the ux and it kind of formalized the the way it should happen and it made it dramatically more visible right like today you know the number of patents is insanely high and like how many people know what's really there uh or how defensible it is or how do you even get it great so you have a patent you want to use it you want to pay the innovators but like how do you how do you do it like if you were to i don't know wave a magic wand knowing what you know as a great connector of fields and knowing like the potential and and the kind of like the reasons what patents are there in terms of funding the innovation custom like really helping the the kind of propagate some of the returns and rewards all the way upstream like i don't know what are like sort of like a if we were kind of redesigning these in crypto how would you how would you what do what do you want in that marketplace okay i'll say three very brief things first thing is um you know the folks at molecule have thought really deeply about this um and they definitely will have better ideas than i will come up here with on the fly second um i want to point to another uh good friend like a it's a big crime maxing friend and he's like you know pretty bullish on not having any ip whatsoever at all so he's basically saying that you know rather than trying to and like um redo the ip system what you should just have is you know basically people are creating companies much sooner so whenever you think that it's like still maybe research project whatever just like strap a company around it um make it a company immediately uh and and try to bring it out in the real world that is obviously um more a longer term game and require some other innovations around it but it's an interesting you know at least like i think food for thought then a third thing that really just came to my mind now maybe someone already has like i thought about this much deeper than i know but you know i think to the extent that you don't really know which of these are useful or not like one way that was interesting to the bit that i just showed about prediction markets um or about forecasting platforms is that they're already being proposed to be used for solving the replication crisis so basically what you'd have is you can predict which paper um would get replicated versus not and then you only try to replicate those that get a low prediction store of a score of like that they have probably low reproducibility and so you can kind of save a bunch of money in time uh by using prediction markets as a kind of like front layer maybe you can also you know predict kind of the use future use values of patterns um and and and and r.i.p in that regard and really try to figure out which are the really useful ones to at least kind of like clear up some of the clutter and then figure out how to how to innovate within them but i'm sure someone has thought has thought about this better than i have but but yeah it is it is certainly kind of like still at the moment necessary able to some extent um and i hear you're on like the sharing of ideas you know i think that to the extent that you can work people have people work toward larger goals um you can just unlock so much more on the long run and and and and it's it's we're really stifled by this kind of like still zero some dynamics that um that are obviously incentivized for them you know it's not like uh they are deeply uncollaborative and these are the most collaborative people if you yeah let's jump to to the topics and so going from um you know biotech neuro attack uh molecular precise molecular machines and then um human collective interaction like um yeah what was the name you used it was a human to human cooperation yeah yeah um the so for each one of these like um can you help us paint a picture because all of these projects are super amazing right like you see some of these and you're like wow that's that's gonna change so much and um we've seen a lot of amazing videos or not just videos but like amazing results um over time and then it taking you know 20 the 10 20 30 40 years um a lot of times that's uh kind of not necessarily necessary but like it's it's kind of proceeding as fast as it could in other cases it's just kind of bottlenecked and the change could actually happen much faster it's just that maybe people like Aaron connecting what based on what you know and what you think about the field um for each of these kind of like let's go through biotech first uh what do you think are like the possibilities in like in this time frame and you know whether it's five 10 15 20 years like what are the kinds of things that we'll you expect will start shifting based on some of the projects that you're looking at i mean for biotech maybe it's it's more like um so first of all i want to say for example i think michael levin at the time when we invited him to our biotech salon to speak about like the zenobot project he really wasn't aware that you know this regenerative component that he has is really related to the longevity rejuvenation industry very much at all um and then we had two of his uh postdocs as our fellows afterwards because i think that sometimes you know like even that the awareness may not may not be immediately there um and i think for okay for biotech i want to name like kind of like just an overlying challenge that i think is blocking the entire industry which is uh really the FDA to the extent that um you know the like the FDA i think is it's doing important work um to some degree but i think that you know you really not measuring what like all the potential life that could have been lives that could have been saved if only we had like faster progress in that area i do want to say that there's a ton of work right now also to creating a little bit more a little bit less friction on that regard but it is still like really one of the main roadblocks that comes up almost in every single discussion in the longevity rejuvenation industry which is the fact that it's really really sticky and slow to get anything through um there are now especially you kind of coming from the crypto space and from the entire corporation group kind of new solutions so you know Balaji obviously talked about a network state with a better FDA um you have even the first charter city and prosper are having like a new approach to an FDA you know how people coming out i think tom bell is doing currently work on surveying different jurisdictions for that have just more competitive governance frameworks around drug development Singapore for example is pretty good surprisingly and they work pretty closely with the government and um and brian kennedy was a buck before us now has moved over there to do much of his long journey drug development so i think really here i have to name a political one which is the like the just the amount of effort time and energy it takes to move anything through the FDA is making it prohibitively expensive for many of the things that could already happen to exist in the real world so and it's killing so many people right like so people don't don't think about the humans that die because the technology didn't get developed like there's this weird kind of moral shift that happens when um hey if it happens in the future it sort of like doesn't count morally weirdly right and um uh we'll hear uh from nick later on about this uh this question but um the from my perspective on the on the FDA timelines like there's probably millions to tens of millions of people potentially more that that you know have died in the last uh 20 to 50 years because those timelines are so slow um and if those timelines were faster all of those people would be alive and i don't know what the numbers might be like that would be a super interesting exercise to do is like figure out like which drugs could we have gotten way faster or which techniques could we have gotten faster and so what is like the total human cost of of the timeline because my sense is that it is orders of magnitude off like in order to kind of do much safer studies um and uh save a few lives now where accidentally killing millions downstream yeah there's actually work on that by i think his name is andrew scott he is at the london business school he's an economist and so he's trying to put basically human longevity and human aging into the quality framework like quality adjusted life years framework to just get like a public health kind of like eye on the on what we're losing by not focusing on the diseases that come with old age earlier on and so that we can at least kind of like compare at work currently are apples and and pears of like you know you have dialysis on one end but you could have really prevented much more of that if you had focused on um you know longevity technologies earlier on so he's trying to put some numbers to that but it's so early still um but yeah but people are working on it yeah let's maybe shift to neuro tech um and this is one of the areas where i wish a lot of this stuff was coming much faster especially given ai timelines um what do you think is possible in the in the near to mid term like you know 5 10 15 20 years what do you think are the sorts of things that are going to start happening um you know we definitely have you know things like uh neural link already kind of very publicly going after this there's a few other super interesting bio uh some bci companies uh they're still kind of like in stealth mode um what a um what do you think is going to be possible like two to two to five years out or or maybe 10 um yeah well i mean i think one interesting bit in neuro tech also that you know is maybe live with cyber as a platform just um brain grid interfaces is really also brain preservation and like brain bio preservation um and so like that is more perhaps more in the longevity area but like you know four things like uh good uh brain storage but also for things like further our prionics and so forth we need to get much much much better at preserving the brain if we're worried about people around us right now dying we need to speed up there and i'm pointing this out here because there's actually a lot of progress in that area like um and it's really kind of encouraging to see but at the same time it's still a pretty fringe technology but i think we could get progressively better at this um so uh we had uh Jean Ebert um he's at Albert Einstein School of Medicine he actually showed that something like brain cell replacement is already possible to some extent with some memory retention um on a very very low level but if you could do that all the time uh you can really already speed things up on a much better level there obviously then i think one thing we really need to figure out is just like much better brain mapping over time in addition to the tools by at Borden that i showed and there's also a wonderful kind of company that i think got started as an FRO who actually Adam who you hear from uh E11 that uh is really really really great at like just like slurping the brain and uh to map it better so i think like actually just understanding like we're gonna get a much better sense of like really understanding what's what's actually happening in different areas that is the number one thing we have to do in the next two to five years and if we can get better at some of the preservation and vitrification i would be absolutely delighted um uh because i think in terms of longevity technologies at least that is something that we absolutely have to get better by and then on the very very long end uh we are having and i'm sure Nick will talk about this but we're having a workshop early next year uh given the short timelines for AGI with Andrew Sandberg from FHI who's done um he basically he wrote the whole brain and relation roadmap i think it was in 2007 and so it needs an update um safe to say a lot new technologies have come online since then and so what we're doing early in Q1 next year is basically having a workshop on whole brain and relations like basically what has changed in terms of like early capability development um how do the pieces fit together and pushing you what if anything does it mean for AI safety because as you are saying timelines have shortened for many many people and maybe hope when the relations are coming back on the kind of like forefront of people again to like actually do something about it but you know there is this long-term goal i will probably hopefully be able to say more about that after we've had the workshop but i'm sure that there's like lots of technologies right now in the pipeline that um you know that that are pretty interesting to uh to bring up again yeah um i'm glad you mentioned cryo um the cryo preservation in my view is one again one of these very deep moral harms that has been done where um it's just sort of like you know for a good reason at the time um there was kind of like a like a backlash to the first few um companies that were starting to to do it and that caused like this huge chilling effect across the entire industry and i think we've basically been set back by 30 to 40 years um and we could have been living it's one of those cases where my guess is that we could be living in in a world where cryo preservation is like an available technique and yet we aren't because because of that stigma um do you think that that's like easing up now like it certainly the overton window has opened up a lot we're starting to see it in sci-fi i mean it was definitely like in the you know in the three body problem i was you know captain america there's like this awesome optimistic view of captain america being frozen and then coming back and so on um and like that's becoming much more much more okay uh do you think that's like definitely like it's warming up or or do you think it's uh um not uh it's still kind of uh you know still like this chilling effect uh you know we're frozen cryo hopefully uh chronix is not warming up hopefully it's staying nice and cool but to your point this is something that i think is often really interesting to look at so this is from an x2p magazine conference in 1994 it's different events that are predicted for example by benford bridge draxler miller is mark miller nicks abo here as well as predicting um and max more he's leading our core which is one of the main cryo organizations now but that mind you is a prediction from 1994 and people are predicting cryonics industry revenues of a billion per year in these timelines now 2022 2020 uh 2015 so you have like very very very different timelines then we actually ended up with which you know i mean we have him on many individual parts that have been predicted here and you know we can get more to individual predictions if you want um but just to say just how optimistic people were back in the days right and do you think that's do you think that technology how fast do you think that technology has developed sorry um these predictions do you think they were off because we just didn't have the tech or do you think they were off because we just didn't work on it the way that we should have uh well i mean like it's i mean we didn't get the tech and much of that is because so for example in most of these fields what you had is like an early spring summer then a crazy winter uh and now it's spring again you know right so for example molecular nanotechnology people like that's the kind of like one tech that we were founded on right that was the early technology that was really kind of discussed in engines of creation and then you know there was a small e-director debate um then there was like a lot of public fear about great room risk like basically a lot of like everything they could have kind of like publicly really been misconstructed there had been then a few people were working on something that they called nanotech which wasn't actually molecular nanotechnology and so we had to re-brand the name from nanotech again to molecular nanotechnology and so all of that happened and that's not only happened in nanotech that has happened across the board and technologies like if you look at ai for example i think early notions of ai always had this more general notion of artificial general intelligence but then when folks started using ai also more for the basically anything in any slide presentation ever um you had to re-brand again to this more long-term vision of artificial general intelligence and frankly like you know you also had lots of different new modalities of actually creating a di that weren't just like logics but but you know we're more like neural net based that that we currently know that it had to be created over time but again you here you had like an early spring and like a winter and now you have spring again you know recently i think it lead around on a few others really brought out this do you think we're past that freezing so um meaning today thanks to the internet and thanks to i think in great part social media we now have a much more decentralized structure to conversations and so the overton window might be shut in like in some mainstream areas but it's very very open with like hundreds of thousands to millions of people um on a lot of these topics do you think like we're kind of like past the this kind of thigh um this kind of like freezing again um or do you think um a lot of these are so like so kind of um not discussed in a mainstream way that uh that it still will be a long time um well it's a great question it all depends honestly it depends somewhat on how crypto is doing yeah yeah i think it's a good indicator like crypto and the interest rates are like the indicators or whether or not we can get fast technology or not it's it's true like it's really true so definitely definitely my my optimism varies by really crypto market prices most um also because people really encryptor like one thing that's so great about it is that people are have already been drawn to the sector the ones that you know that were early have been drawn to it when um it really wasn't very much yet and so they already probably have a pretty open mind and so it means that you know oftentimes people especially in uh you know at conferences like yours like they're able to grasp these segments and they're able to like already have like the overton window it's it's only a little bit wider like if they even release this one um and so and so I think that you know to the extent that it's just like many of these technologies have sped up significantly since crypto has been a thing because many people also put a lot of resources to those areas that are encryptor that you just didn't see on that scale before and it kind of has become cool again so I definitely do see it like I mean like five years ago and many of our five years ago at our longevity conferences there were almost no investors now when you go it's an entirely like VC um industry has spun up around longevity investing I hope we get the same to molecular and nanotechnology uh in in five years um to bring a bit in phases it's already happening to some extent but one thing that people are always really really cautious about is so many people have been burned that were early uh that people have become you know kind of so kind of like a little bit more cautious um a little bit like there's more discussion about the language that's supposed to be used there's more discussion about also like oh that person over there is a little bit too out in the open about discussing you know very ambitious tech and we should really just not feed that to the public yet and so it's really difficult to kind of coordinate just even on like you know what how far out do we get uh you know and and how and how how much do we how how much do we take the the traditional um main scene with us versus like really actually just pushing at the frontier it's difficult yeah I think that opening over to windows is definitely difficult but it's a skill that can be learned in practice you know we should get elan to teach a master class on this because the overton window opening on bci in the last few years has been astonishing you know I think five years ago uh four years ago you couldn't at all suggest the idea of like putting wires into people's people's brains and now like uh varies you can sort of like go back and watch um things that elan and other people were talking about and in a matter of like two years by talking about specific things you can kind of see like quarter over quarter the adjustment that happened and the total opening of the of the of the window to the point where like now there's like it's sort of like a mainstream topic that can be discussed um and I think that's that's uh yeah it's pretty it's pretty awesome um let's maybe jump to uh sorry go ahead no I just want to say one thing that really pains me is that it's too late with quironics like the thing like people always talk about uh like you reach out to me about quironics when it's too late when their loved one has already passed and then they suddenly realized it was like oh actually I would have wanted that safety net even if it's like really really unlikely to work and so it's like the same with privacy you only realize how important it is once it's too late and so I encourage people to think about these things early as they are still uncomfortable but yeah as they're still possible on molecular machines um that was uh pretty cool to see the the the minecraft uh interface that director was talking about was showing and it was also really cool to see um just kind of like using proteins and so if I understood correctly it was like using proteins to kind of assemble the the the machines uh how far like especially with things like you know alpha fold being able to you know solve the protein folding over over winter break or like that was amazing that was such a great result um how much do you think we're gonna be able to pick up the speed on this right because like it could be we could see this kind of series of breakthroughs over two four eight maybe ten years into getting us like some extremely powerful tech um do you think that's like viable or do you think it's still kind of like you know maybe 15 20 years out to get to or longer um yeah it's a great question I think one thing and people in the field really don't agree like um you know and so I think one thing that needs to have like so what you see is breakthroughs in individual parts so for example DNA origami had like absolute like an insane one year uh this year like a will she and like a few others have produced just really like lattices of DNA origami um and you know you like they can like literally I mean it's maybe I'll I'll can bring up some some of the images that they've been already like produced with the with the in a origami but like that's great um putting folding amazing and like you know you have great work on that on that end I think some of the things that you know at least the Drexler like very ambitious system of molecular analysis what they're hoping to do is something that goes beyond uh what's currently possible within a lab and that is really building very complex machinery it's really like building complex like materials it's building complex um uh you know like complex motors uh and and really like rebuilding much of the physical infrastructure and then creating new physical infrastructure new physical uh components uh from scratch and so for that and many of the systems that are currently where Pro is speeding up they need to be integrated and and that is not something that individual researchers in one lab have the time energy funding to do and so that really requires a lot of kind of work on even simulating how these systems could function which is why I think the work that he was showing is so interesting and why much of the work that I've been showing was more on the simulation end and on the design end because we don't really know at that level we don't have a good intuition for how do these systems integrate with each other how does it do they break if we slot individual pieces together can we just like you know how creative can we get and so the better we are able to simulate that the more we are able to figure out the the the last mistakes we have to make basically in building these systems and mistakes are very costly and take a lot of time and so I think the simulation arm and like and that's why also the kind of like physical simulation that I showed you know like of like people actually using a whole building as a computer and being able to move like molecules around and that's so cool because you can actually develop a better intuition for what's actually going on at that level which is something that we currently don't have so I think the more we are able to get better on that end the faster I am or the more hope I am on the progress long-term in building these very very complex systems yeah and on human human cooperation I think that there the timelines can thankfully be much faster I mean we've seen tremendous work through through their crypto space in the last five ten years um uh what are you sort of like excited about it seems to me that if we get things right we could have crypto networks funding R&D at the scale of nation states in five ten years and another sounds crazy but the budget of the you know the US budget funding science and tech is something like 60 to 80 billion a year or something like that and something that's I think like the you know NSFNAH and others and DOE and whatnot that seems like something that crypto networks could achieve in ten years do you think that's like viable or what are the sorts of things that you're excited about yeah I'm definitely yeah very very excited about that I mean I think that in general cooperation you know is like is restrained by search costs you know it's restrained by commitment costs and is restrained really by you know enforcement costs and you know there's like few transaction costs maybe that are beyond that but I think you know one thing that Mark Miller always said and who I've been like writing this book with is like you know whenever you lower the costs of cooperation by like a lot then it's almost like a shift in quality and so you know like the internet like what it gave us is this like incredible ability to cooperate with strangers across the world but the interactions that we have are still pretty limited to the extent that you know we can like exchange information we can sometimes exchange like you know what we can definitely do financial transactions but like the kind of like complex contracting that you know the complex interactions that we do with people around us that we trust highly and that is restrained like and we cannot do that with everyone in the world even though there's probably like a bunch of different kind of like interesting collaboration opportunities that we would have so I think I'm genuinely also really bullish on the kind of like general space of smart contracting and like what can that allow like right one thing I didn't talk about is like kind of like the different types of contracting tools that people are already experimenting with like you can have something like split contract at the beginning where like part of a contract can be automated part of a contract can be proposed so you know we don't want to automate everything up front maybe we want to have something that is still left up to human arbitrators through like and I know Kleros or like some human arbitration system and you can kind of like iterate from there and at the same time like you know you have dominant issuance contracts you have all of these like different contract experiments that I think are just like really wonderful and I think this kind of like ability to really unleash the power of cooperation across the board is something that that I think I just don't see very much very much like excitement around it in the like traditional mainstream or something but that I think could just be so influential and then another thing that you know we I had touched on briefly that I think will be really interesting is this intersection between crypto and AI and so you know like you know you have this notion that crypto is more decentralized and AI is more centralized like that doesn't I think always square perfectly like and it could be really interesting to like just really look at what are like interesting tools in machine learning that and interesting applications that you can only unlock because you have a cryptographic element to that and so I showed a bit about this work from your discussions on differential privacy and trust from open mind has done really incredible work on federated learning there's a bunch of zero knowledge enabled machine learning there's like peer-to-peer intelligence markets and and so there's all of these things in which kind of like crypto can bring this kind of aspect of security and privacy to machine learning that we do not see often in traditional AI or it's not like maybe by default and I think that's something extremely interesting not only because I think that you know like centralized AIs that you know have access to a ton of data are like potentially worrying but also because I think that you can just unlock entirely novel applications like McConny talks about AI DAOs and has a really interesting series of blog posts they probably come with their own with their own risks really but some of them can already be created like by just you know using existing existing GP3 and so forth that's already on the internet you could always spin a DAO out of that it's like it's kind of like a few things that that I think are like leading naturally to like interesting mixes where you just yeah mix the two fields a bit even though they look in a natural position to each other so I think those are at least a few intersections where I'm relatively interested where progress will arise. All right I want to get to Q&A with with the audience so I'm going to ask everybody who is here present to just use the mic in the up here so if you have a question just stand up and walk up while people are doing that and please form a queue so that that way we can go from question to question to question. Well I'll ask two more questions before that so first one a hopefully easy one how when you think about AGI how far away do you think it is just based on everything that has happened in the last five ten years if you were to guess what's your current estimate on on the timing maybe like where's your like probability mass in the in your estimate. Yeah really hard I don't I'm not sure if I can give a good answer to that I think that's a great question for for Nick also and Max to ask I think in the broad sense you know there are different schools within AGI and also within AGI and there's certainly the more kind of like the traditional school that also superintelligence the book unleashed of you know potentially like more decentralized intelligence and then there's the comprehensive AI services model that is basically what would it mean if technological development was like progressing faster and faster and able by AI but it looked a little bit more like a decentralized kind of like ecosystem of services that are kind of like in broad cooperation with each other and so it really also depends a bit on like and like the type of view that one subscribes to there has recently been a very interesting debate between Eleazar Kowski who has a very very very short timelines at MUI who's more on like I guess perhaps I don't want to paraphrase people like perhaps on a more of a centralized AGI approach and then Paul Cristiano who basically just advanced was advancing the idea that like progress can perhaps just speed up in a little bit more of a kind of like distributed or like at least sorry not distributed but like at least piece me away so you have a slower takeoff at least at the beginning and they even thought it was really difficult to agree on exact at timelines so basically they were trying to make a prediction and and I think the only thing that they could agree on that could potentially be different was whether or not I'm hoping I'm getting this right now you should definitely look this up when an AI would win a specific or would outcompet a human in a specific math problem I think in a math Olympian and by 20 27 or something I'm not entirely sure about the year but basically it's it's really difficult I think to get your timelines right because even if you have very fast timelines you would still think the timelines look very similar to someone was short with long timelines until one day they don't and so this kind of compounding effect in computing is just something that makes it really really hard so I would leave it to the experts there to to make a prediction and and there's certainly lots of discussion out there right now and many of the wonderful AI blocks that are out there great thank you actually just jump straight into into questions I'll save mine for if there's a gap first question and if you can say your name for for the record hey Alison it's Danny O'Brien from FFDW hi so I'm the designated questioner for today and I guess I'm going to give the default question that I think you get asked a lot but I would be really interested in your kind of default reply which is I think when faced with these sort of existential threats and opportunities a lot a lot of people particularly in the very sort of pessimistic period that we're going through at least in the kind of the mass media often their responses to call for a caution right so they'll they'll people will talk about things like the the precautionary principle and this idea of holding back because we need time to think through all of the consequences of these actions I've always seen the Forsyth Institute as being kind of at the very opposite range of responses to the precautionary principle but I wonder how when people are calling for those kind of actions how you how you respond to to to that kind of concern thank you yeah I don't want to like straw man anyone's opinion here so I think like in general and hi Danny it's nice to see you but in general I think you know one thing that we also have to think about is the cost of not developing technologies and not of not developing the progress that comes with them and so I think that you know a on the long run we won't be able to survive on earth if we don't develop technologies because eventually the like sun will boil away much of really much of the surface of the earth as far as we can currently predict right so eventually we have to if you have long timelines I have to do something about that and so I think that we need tech progress and then the second thing is that you know you have risks speeding up I think natural risks also that we have to learn how to deal with right like many of the things that are currently bits and pieces that are that we take for granted in hindsight you may really think about as like you know very big more atrocities that have happened so I think that you know there's definitely a cause for the non-development of technologies as well and so I think that we need to we need to build kind of like a sober pragmatic risk assessment into our technology development as we go forward I think one thing that we're trying to really do at Fawcett is to really kind of tease out with the technologies that are working on these areas like what are areas that you think can can can need to potential risks here and like are there specific technologies that we can speed up that we can speed up in their state I think there's also a few technologies that are just really like generally pretty good to have like computer security for example like much of that tag that seems to me almost like a no-brainer that like we should really really push ahead on this like there are certain people a few people at least that think that there could also be risks from that but I think that you know that to the extent that there's a few technologies that we just have such a large upside I think that we should focus on on speeding those up I do also just want to point to the uncomfortable truth that there are you know just generally race dynamics at work in the world and so I think that to the extent that you really think that you have a like good approach to technological risks there is also something to be said of like building a secure infrastructure for them and building like a you know a beneficial I think application system from technologies up front so that that can take foothold in the world and before perhaps like more you know and like totalitarian use cases or so forth can grab a can grab a hold so yeah there's lots of individual bits and pieces floating about in my head but I think in general yeah there is something to be said I think of having like a sort of like pragmatic risk approach in your technology development and it's a difficult trade-off to have for sure. One question from Twitter first Dave asks Allison you said a lot of academics already have a company they just don't know it yet what if universities created a legal business entity or company for every student by default or the ASS could extend to labs or projects could this be put in chain what are your thoughts on that? Oh that is a great question um I so sorry they have the technology to have a company um like I think that the mindset really is really such that they may not be a really great CEO for their company and many of them don't want to many of them want to stay in academia so what often happens is that they could become a scientific advisor to someone that then becomes a CEO of their company and so there's definitely it's a little bit more complicated than you know really just having the idea for something but one thing that we often get kind of like or got at least in the in the midst of that we are like definitely still half on our kind of like things to do for late this year early next year is that many of our for example much of my day-to-day work is matchmaking between our fellows that are either in academia thinking about spinning out or that are already in a company early stage company our founders on the one hand and then between funders that are our donors or like investors philanthropists on the other hand and so this matchmaking was pretty tedious and so a few folks basically like were like why don't you just start like a non-profit venture fund and so the idea of that would be that the carry of such a fund and the management fee would flow back into the non-profit rather than go into any individual and that through that kind of like through that funding we could create further prizes and fellowships down the line meanwhile investors would still be getting would still be getting profits on what they invested in but just the carriage the carrying the management fee and so there's definitely been like these kind of like different approaches by which you know we were maybe getting a little bit more organized by just pulling on these technologies and pulling them more into our ecosystem because I think one thing that academia isn't doing very well yet is creating a kind of like tech friendly ecosystem so I don't actually know if like universities gave you a company wrapper if that would be good enough like I think you need much much more you need like people that are really just go get us and can like you know push like push companies out of the ground and that is a very different skill than you get taught in academia so I think it still requires this ecosystem that that we currently do by matchmaking but hopefully you know over time there are just like better vehicles available for this all right go ahead and hey my name is Mike McCoy it's actually my first day here at protocol lab so I hope it's not a weird question but within privacy enhanced technologies or privacy preserving technologies have you ever found like economic models or specific data points that are just certainly more valuable than others to actually share and create those type of constructs to share with within health care and life science partners like are there certain ones that stand out more than others that you have just seen that markets and people and and life science companies especially researchers would want to be able to find out more of and that it's worth to create like the randomness the fuzziness of that data at scale to then share for the insights of it like have you found any details there um well there are certainly in health care I mean like the one that I showed was from George's cases at open mind and usually you know people bring up health care is like a good use case for better ways of dealing of better ways in which privacy preserving technologies could be used um for to unlock a variety of different um different new applications like on the one hand health tech is one of these areas where like it is just the most sensitive data because oftentimes your health data is not only data about you but also about your family um and so it is a really really tricky problem actually um and you know we shouldn't let our guards down too much by the tech that we have I think at the moment but on the other hand some of the applications really that you could do like for example some of the machine learning bits that we've seen in the first slide that I showed about creating better aging clocks like currently often the like n is one in much of the longevity applications because people do some like you know experimentation on themselves they you know perhaps like you know really track like Brian from Brian Johnson from Kernel has this really insane effort and called the blueprint where he's like really tracking uh really well most of his biomarkers and just exercises a ton eat super healthily and shares all of that but that's n of one like if we could do that across everyone who has like ura rings who like you know uses most of the most of the aging clocks that are currently already out there and perhaps there's some of the interventions um and that are that are currently already out there and and was able to actually cross-compare how that fared with different health profiles that would be a really great use case it's really difficult on the one hand people don't often trust people to measure their own uh their own um their own biomarkers well like you know like people just don't trust kind of like self experimenters to have a good good check on this but you could maybe also solve for that but I think so I think it is the really really interesting application area there's a few federated learning uh approaches that I know of a few different privacy approaches that I know of um and I think across those we'll be able to unlock a few of those silos just in a better way hopefully uh but yeah we're not there yet one more question from Twitter uh for some of those radical societal ideas how do you think about building a new system versus reforming an existing system um and maybe I would add like what are ways of trading off the costs and benefits there yeah um well I think that you know to the extent that much of the you know our crypto and like human AI cooperation group they're really working on the building new systems uh part uh you know that some of the work that I've shown about like you know your presentation problem and then Balaji's network states those are certainly like just new ways of doing things that we currently do um at the same time you know some of these technologies are also like they just need a bit of funding um you know they they they really could run with existing tools they maybe need one or two connections over here over there and a bit of funding and uh and then off we go so it really depends a bit on it depends on the on the scheme and on the technology at hand I think that um the great thing about building new systems is the multiplier facts that you can have like we've learned a lot through the last hundred years of like what works and what doesn't work for innovation and so to the extent that you create entirely new like protocols right and like which you can run things you can really unlock um ecosystems in a much more foundational way than you can by having within the current constraints on the other hand for some other things like especially in the longevity industry leads to my mind they're so urgent that we just have to use what we've got and just have to plug away at it um and you know that was we funding constraint which it still is but not so much anymore so it's uh different fields require different things like longevity I think really funding funding funding more AI research moving in then um newer technologies definitely requires better tool development uh we're not really quite there yet but I think we're really getting there molecular machines there I think the integration of individual workflows is really important to the extent that we can do that in the existing system great uh and then you know the kind of a human computer um uh cooperation I think you know they are already doing really wonderful things and I think that that is the more experimental space uh that perhaps I'm less worried about but here I can just also point to the more privacy preserving machinery bits as exciting areas to focus on more so it depends on the space I think but it's a great question so thank you for the presentation this has been very interesting uh I'm Sean from protocol labs uh given the business cycle we're at now spring summer kind of fall what is the research approach that you see being most effective by researchers who have gone through previous falls previous cycles of disillusionment for say uh cryogenics or other fields how can they continue to make long-term progress on in this area over decades as the business climate changes yeah great question um is there some talent not too many fields you actually have like oftentimes it requires it is oftentimes it is new people stepping in getting excited not remembering how difficult and um heartbreaking it was to see their field be discredited and destroyed um then they're meeting old like the kind of old old g's in the fields um that have had had to go through some of that and then them having a really good collaboration uh and then it becoming like you know a new project or a mix of a project um and so I think you know there's certainly a ton of people that like I I'm always surprised but when I speak to folks from you know from for the early days their ideas are as relevant now as they were back then like Mark gave a great talk on the computer security as the future of law uh that was like 30 years ago or something him and example were debating virtual societies um in that talk um and much of the smart uh kind of infrastructure that people um that people are now building out and I sometimes I asked him like aren't you bitter that um uh you know that some of that stuff um you know really took so long and that now there's this like uh that field is like kind of like growing but there's like so many new projects in that space and he's like I am so delighted you have no idea it is the best thing I could have hoped to see and so I think it is rewarding for many people that I certainly interact with that were early on to finally see new spring and new people moving into their areas and just seeing the things that like their babies you know uh their babies that were like destroyed and discredited for a while to see new people breathe new life into them and so that's really really great and to the extent that I think uh those folks often have a lot of to learn from people that were early in this space because they've no talk about the mistakes that were made and many of the ideas are still so so relevant um so I think that like collaboration between people that were early and people that come new uh that know uh that are not that you know burned out yet um that I think is what I'm like super excited about and at the same time I'm always yeah really uh by how relevant many of the early ideas that I still sometimes find in our archives still are like some of the stuff is is still future thank you right I'll ask the last question um Allison you've guided us through a really wonderful vision of um of how a lot of these technologies could come in um and change a lot of facets of our life like everything um the things that we're talking about could reshape a lot of facets of our society and with the risks in mind and with the with the potentials for problems and so in mind um can you talk to us maybe about like your super optimistic vision for the future uh like what you know suppose that we succeed in handling all of those risks um what is the what is like a a really great future that you kind of think out and you know this could be like short term or very long term um what are you sort of like hopeful for and um aspire to uh yeah well I could um you know really I think go through a few of them that we've discussed here but I think um I think ultimately I'm really excited about um this concept that I've also discussed recently with Anders again that certainly doesn't come from my end but like the concept of morphological freedom um and so the idea that you know currently there are so many ways in which we'd like to grow in which we'd like to expand in which we'd like to connect with each other that uh we are being just artificially constrained by and I think you know our normal way of thinking about kind of life is in this more existential list mindset of like you know we take our constraints as given and we try to create as much meaning as we can within the given constraints but you know I think this kind of like more vision of existential hope is asking the opposite question of like what would it mean if instead of like lowering our expectation to fit the current realities what if we could like improve on reality to meet these hopes and dreams that we have or at least once had when we were kids before we forgot about them and and what would it mean to actually reshape reality to be able to meet these hopes and dreams because I think that you know back then there were things that I barely even remember and I can barely conceptualize because you just kind of like get boggled down by kind of like the default reality in so many ways but it often just only takes kind of like looking at that field again and asking what could happen and getting a few people excited about it um to kind of like make that progress so I think that ultimately really like creating that culture that shift from from lowering our expectations to meet the default reality to like lifting default reality to meet our hopes and dreams and like to really have give people the morphological freedom to be the people that they'd like to be and not if they don't want to be that um that I think is what I'm ultimately super excited about and and and I think that yeah really the the sky is probably the then even not the limit but but yeah this kind of like notion of freedom and human ingenuity and and maybe also other intelligences in ingenuity is I think something that's very inspirational to me. Well thank you thank you for all of the work that you do it is a great you know delight and honor to work with you throughout all of these you know key fields and um yeah thank you for writing the book the book is phenomenal for my point of view it is um uh an amazing work of of proper philosophy like when we think about philosophy and like the we look through history as like books that take in what we know at any given point in time and try to help it make sense and paint a picture of like how we should be living our lives and how we should be thinking about the future and and what we should we aspire to that is like what philosophy um originally was about and it is like truly one of the few works these days that can be called proper philosophy that like actually take in all of the scientific advancements that have happened and the potential scientific and technological breakthroughs and truly gives us a picture into how might we navigate this this the set of possibilities and gives us a very practical down-to-earth guide of like hey do these things and you will we can as a species continue to play the civilization game so thank you so much and thank you so much for joining us today take care I do have to just give you a very very big detail I think that you know it is because of individuals like you that many of these things are now happening and they're like more than philosophy now and so I think your protocol apps and popcorn and so forth like it is thanks to you know much of that ecosystem that you guys are creating that now these things that people have been dreaming up for a long time can finally like get closer to reality so thanks a ton thank you for having me and yeah it's a really great community that you guys are building and I'm really looking forward to much of the work that's coming out of it thank you so much all right thanks everyone and see you tomorrow with nick nick nick mushroom